Simultaneous measurement under resting-state of autonomic nervous activity and brain activity by near-infrared spectroscopy alone

Near-infrared spectroscopy (NIRS) is widely used for brain-function imaging, but the raw signal also encompasses various other physiological factors such as heartbeat. In this study, we aimed to extract heartbeat-related signals from the NIRS signal and evaluate autonomic nervous activity and brain activity simultaneously. For that purpose, simultaneous NIRS and electrocardiogram (ECG) datasets were acquired for 10 healthy volunteers, and the heart-rate intervals obtained from each dataset were compared. We also investigated factors which could affect the extrication of heartbeat-related signals from NIRS data, such as the number of NIRS channels, type of NIRS signal, and resampling interval of NIRS data. We found that the heartbeat-related signals could be effectively extracted using total hemoglobin data by averaging all channels (52 channels) after band-pass filtering and upsampling to the same 4 msec (250 Hz) as the ECG sampling interval, which showed a high correlation with ECG R-R intervals. We also found that indicators of autonomic nervous activity could be calculated with NIRS data and had a high correlation with those calculated from ECG data. Despite its simplicity, our method demonstrates the possibility of analyzing brain activity while also considering autonomic nervous activity using NIRS data only.